Compounding and its locus of realization: Evidence from Greek and Turkish

Abstract

This article deals with the locus of realization and the grammatical nature of compounds. First, it suggests that a proper delineation of compounding should be given on formal grounds and that an approach relying on pure semantics is misleading. Second, it proposes that the diversity of views for defining compounding and the variety of theoretical approaches that are put forward for the analysis of compounds are highly dependent on the data under examination. Third, it defends the position that compounding cuts across two grammatical domains, morphology and syntax, assuming that they are distinct structure-building modules. On the basis of their structural properties, compounds can be distinguished into morphological objects and phrasal units bearing an atomic status, depending on the language one deals with. The first category includes compounds resulting from morphological rules (or templates/schemas), and involves units specific to morphology; the second category contains phrasal compounds, which are semi-visible to syntax, but their structure is derived in syntax, in that it is not based on morphologically-proper units and is not the product of morphological rules or templates. Claims and proposals are illustrated with data drawn from two genetically and typologically distinct languages, Modern Greek and Turkish, which significantly diverge as far as their compound formation is concerned.